
According to reports from technology media outlet FundaAI, Intel has reached a cooperation agreement with Taiwan-based UMC, the world’s second-largest pure-play foundry, to jointly develop 12nm and 3nm semiconductor manufacturing processes.
The collaboration marks a potential strategic shift for UMC, which has long focused on mature-node technologies, as it could re-enter advanced semiconductor manufacturing through Intel’s U.S.-based fabs. The move is also seen as a key step in Intel CEO Lip-Bu Tan’s broader strategy to strengthen the company’s foundry business and capture greater market share from TSMC.
UMC withdrew from competition in advanced nodes below 10nm in 2017 and has since concentrated on mature processes and specialty technologies. However, with accelerating demand for advanced chips driven by artificial intelligence, the company’s global foundry share has gradually declined. In recent years, it was overtaken by SMIC, and its market share reportedly fell to 3.9% in Q1 2026, ranking fourth globally.
Meanwhile, Intel has been aggressively reshaping its manufacturing strategy under its “IDM 2.0” framework, first introduced in 2021. The company resumed investment in leading-edge process development and expanded its foundry services ambitions, aiming to become the world’s second-largest foundry by 2030.
To accelerate its foundry expansion, Intel partnered with UMC, which brings extensive customer relationships and experience in high-volume manufacturing. In January 2024, the two companies announced plans to co-develop a 12nm FinFET platform, leveraging Intel’s manufacturing capacity at its Ocotillo campus in Arizona to support joint customer development.
However, Intel’s large-scale capacity expansion efforts and uneven business performance later contributed to financial pressure in 2024. After new leadership changes in 2025, Intel reportedly paused several overseas expansion projects while maintaining focus on advanced-node manufacturing and U.S.-based foundry operations.
Recent updates suggest that progress on the 12nm program is advancing steadily. The process design kit (PDK) is expected to be delivered to customers in 2026, with initial tape-outs planned for early 2027 and mass production targeted for late 2027. The node is expected to serve applications such as Internet of Things (IoT) devices and Wi-Fi connectivity chips.
UMC President Jason Wang confirmed in May 2026 that validation work for the project is currently underway at Intel’s Arizona facility and is expected to be completed by the end of the year, paving the way for production ramp-up in 2027.
Beyond 12nm, the two companies are also exploring collaboration at the 3nm technology level. Under the same partnership logic, Intel would contribute advanced manufacturing capabilities and FinFET process expertise, while UMC would leverage its foundry service experience and customer ecosystem to improve fab utilization.
The proposed 3nm production is expected to take place at Intel’s Ocotillo campus in Arizona, including facilities such as Fab 12, Fab 22, and Fab 32, with some reports indicating Fab 52 as a key site.
For UMC, this collaboration could provide a pathway into leading-edge manufacturing without the need for massive capital investment in EUV lithography systems or standalone advanced-node fabs, significantly reducing financial risk while gaining exposure to next-generation process technology.
Industry reports suggest that although the 3nm collaboration mirrors the structure of the 12nm agreement, its ambition is considerably greater. The two companies are reportedly targeting a process platform competitive with TSMC’s 3nm-class technologies, directly entering the high-end foundry segment dominated by TSMC and Samsung Electronics.
If successful, the partnership could mark UMC’s return to advanced-node development after nearly a decade and represent one of Intel’s most significant steps yet in positioning itself as a credible foundry alternative at the 3nm level. As of publication, neither Intel nor UMC has officially commented on the reports.